Neoadjuvant chemotherapy versus surgery alone
for locally advanced adenocarcinoma
of the stomach and cardia
Final results of the EORTC phase III randomized
trial 40954
C. Schuhmacher, P.M. Schlag, F. Lordick,
W. Hohenberger, J. Heise, C. Haag, S. Gretschel,
M. Mauer, M.P. Lutz, J.R. Siewert
Rational for the EORTC Trial
…at the time of protocol preparation (1998)
Remarkable response rates in clinical trials
of locally advanced and metastatic gastric cancer
Phase II trials were always
criticized for an incomplete pretreatment staging
Postoperative chemotherapy (adjuvant trials)
could not demonstrate significant benefit
Adjuvant therapy had to be often delayed:
postoperative deterioration of performance status
and complications
Background
Multimodal therapy in Gastric Cancer:
Trials performed at a comparable point in time in
!
The Netherlands, Great Britain, Switzerland, Italy, France, USA
Staging
Prospective phase II trial with endoscopic ultrasound and extended
diagnostic laparoscopy: Exact cT-category and detection of peritoneal
carcinosis or occult visceral metastases
Regimen
EORTC 40953: Combination of cisplatinum, folinic acid, and infusional
5-FU (good activity and well tolerated)
J Clin Oncol. 2007;25(18):2580-5.
Background
EORTC 40954
2 cycles (d1 – d49)
• Cisplatin (50 mg/m²)
• 5FU (2000 mg/m²)
• FA (500 mg/m²)
q 2 wks x 3
q week x 6
q week x 6
+ surgery
Does a purely neoadjuvant chemotherapy yield an
overall survival benefit in the treatment of locally
advanced gastric cancer?
Design
Randomize
CS
1 cycle (d1 – d49)
Cisplatin/5FU/FA
Restaging
S
<= 14 days
Resection if
possible
within 3 days before cycle 2
<= 4 weeks
Resection
If no PD/tox/WHO 2
Cycle 2
<= 4 weeks
Resection
Follow-up
Endpoints
Primary:
• Overall Survival
Secondary
• R0 resection rate
• Time to progression
• Toxicity during preoperative chemo
• Postoperative morbidity
• Effect of chemo on lymph node metastasis
Main eligibility criteria

Histologically proven adenocarcinoma of the stomach (∑ AEG Type II
and III)

cT3/4 NX M0 (only exception M1lymph)

No histological confirmation of suspicious peritoneal lesions at
diagnostic laparoscopy

No prior chemotherapy and/or radiotherapy

WHO PS 0-1; 18 ≤ age ≤ 70; adequate organ function
Randomization stratified by:
- institution
- primary tumor cT3 vs. cT4
- localization of the tumor (upper vs. middle - lower 1/3)
- gender
- histological subtype (intestinal vs. non intestinal)
Statistical considerations

Primary end-point: overall survival
282 events required to detect an improvement in median overall
survival from 17 months to 24 months (HR=0.708) with a power of
80%, using a two-sided logrank test with a significance level of 4%
- initially planned to perform a first analysis on patients who had
an extended D2 resection
- next to perform a second analysis on all randomized patients with
a significance level of 2%

N=360 patients (180:180)

4 years accrual

2 year additional follow-up
Recruitment

Date of activation:
July 15th, 1999

Date of closure:
February 10th, 2004

Study was closed early at
144 pts (72:72) because of
poor accrual.
!
N=144/360 (40%)
Patient characteristics (N=144)
CS
(N=72)
S
(N=72)
56 (38 - 70)
58 (26 - 69)
50 (69.4%)
50 (69.4%)
PS 0
48 (66.7%)
55 (76.4%)
PS 1
24 (33.3%)
17 (23.6%)
T3
68 (94.4%)
67 (93.1%)
T4
4 (5.6%)
5 (6.9%)
intestinal
33 (45.8%)
33 (45.8%)
non-intestinal
39 (54.2%)
39 (54.2%)
upper third + cardia II, III
37 (51.4%)
39 (54.2%)
middle third
20 (27.8%)
18 (25.0%)
lower third
15 (20.8%)
15 (20.8%)
Age Median (years)
Sex Male
WHO performance
!
Clinical T stage
Histological subtype
!
Tumor localization
Treatment
S (N=72)
CS (N=72)
Started chemotherapy
n=69 (96%)
1 patient refusal but had surgery
2 patients with no records
!
Underwent resection
n=68 (94%)
Completed
chemotherapy
n=45 (63%)
Discontinued
n=24 (37%)
Underwent resection
n=70 (96%)
Toxicity (n=8),
Patient refusal (n=5),
Other (n=7)
PD (n=4)
2 patients with
unresectable tumors,
1 patient with liver
mets discovered
intraoperatively,
1 patient without record
Reason for discontinuing chemotherapy
Major reason for protocol discontinuation
normal completion
CS (N=69)
45 (65.2%)
PD
4 (5.8%)
toxicity
8 (11.6%)
renal toxicity (max G2)
cardiac toxicity (G3)
nausea and vomiting (max G3)
neutropenia (max G2)
2 pts
1 pt
4 pts
1 pt
refusal
5 (7.2%)
other
7 (10.1%)
worsening of symptoms (not PD):
non-compliance:
judgment of investigator:
infarction of the pons cerebri
(unrelated)
2 pts
1 pt
3 pts
1 pt
Surgery

Resection consisted of a subtotal or total gastrectomy with
extension depending on the localization of the primary tumor with
either a D1 lymphadenectomy (7 patients) or preferably a D2
lymphadenectomy (130 patients).
CS
(N=70)
S
(N=68)
D2 lymphadenectomy
67 (95.7%)
63 (92.6%)
+ transhiatal resection
31 (44.3%)
35 (51.5%)
+ hepatoduodenal lig. & rt. retroperitoneal excision
20 (28.6%)
22 (32.4%)
Subtotal distal resection w/ systemic LN resection
1 (1.4%)
2 (2.9%)
Multivisceral resection
6 (8.6%)
12 (17.6%)
Type of reconstruction
Roux-en-Y
Pouch
48 (68.6%)
17 (24.3%)
50 (73.5%)
12 (17.6%)
Postoperative complications
Postoperative
deaths
Postoperative
complications
CS
S
3/70
(4.3%)
1/68
(1.5%)
2 sepsis
1 cardiac arrest
+ PE
1 sepsis
19/70
(27.1%)
11/68
(16.2%)
Response to chemotherapy
CR
4/69
(5.8%)
PR
21/69 (30.4%)
CR+PR
25/69 (36.2%)
(95% CI: 25.0% - 48.7%)
Pathology
!
CS
(N=72)
S
(N=72)
R0
59 (81.9%)
48 (66.7%)
R1
9 (12.5%)
15 (20.8%)
R2
2 (2.8%)
5 (6.9%)
Not operated
2 (2.8%)
4 (5.6%)
• R0 resection rate was significantly higher in the CS arm (P=0.036)
(according to the pathology report)
• By intraoperative assessment, R0 resection was achieved for 63
patients in each arm (87.5%)
Pathology
!
!
CS
(N=70)
S
(N=68)
Median tumor thickness (mm)
11.0
13.0
Median tumor length (mm)
50.0
57.5
Median tumor width (mm)
40.0
45.0
P-value
Extent of tumor*
T0/1/2
T3/4
46 (65.7%)
24 (34.3%)
Nodal status*
N0
N1-3
27 (38.6%)
43 (61.4%)
13 (19.1%)
52 (76.5%)
0.018
no lymphangiosis
41 (58.6%)
23 (33.8%)
0.011
*2 unknown, 1 missing
34 (50.0%)
31 (45.6%)
0.113
Survival
100
90
80
70
60
!
50
40
30
20
Overall Logrank test: p=0.466
10
0
(years)
0
O N
35 72
32 72
1
2
3
Number of patients at risk :
58
48
34
61
49
41
4
20
29
5
11
15
6
4
6
7
Treatment
S
CS
Survival
CS
(N=72)
4.7 years
Follow-up
Treatment
CS
S
Patients
(N)
Observed
Events
(O)
72
72
S
(N=72)
4.1 years
P-value
0.637
Hazard Ratio
(95% CI)
P-Value
(Log-Rank)
% at 2 Year(s)
(95% CI)
32
0.84
(0.52,1.35)
0.466
72.75
(60.66, 81.67)
35
1.00
67 events: power of 25%!
69.92
(57.68, 79.24)
Progression-free survival
100
90
80
70
60
50
40
30
20
Overall Logrank test: p=0.200
10
0
(years)
0
O N
44 72
40 72
1
2
3
Number of patients at risk :
44
34
28
56
41
31
4
16
24
5
11
13
6
4
5
7
Treatment
S
CS
Progression free survival
Progression documented
Patients
(N)
Observed
Events
(O)
CS
72
S
72
Treatment
CS
(N=72)
34 (47.2%)
S
(N=72)
43 (59.7%)
Hazard Ratio
(95% CI)
P-Value
(Log-Rank)
% at 2 Year(s)
(95% CI)
40
0.76
(0.49,1.16)
0.2003
58.57
(46.16,69.07)
44
1.00
47.89
(35.93,58.88)
Time to progression: HR=0.66 (95% CI:0.42-1.03), p=0.065
but two more deaths due to postoperative complications in the CS arm
Discussion based on the results
Accrual unexpectedly low
Competing trials on the market
High number of proximal tumors
Epidemiological evolution
Unexpectedly high cT-category
staging error
Cardia is covered with fat, not
serosa (“T2b”)
Chemotherapy with only 63%
complete courses
Toxicity of PLF still too high
Pathological findings as expected
New effect on lymphangiosis
Surgery arm with a median
survival of at least 36 months
(better than expected)
Effect of D2-Lymphadenectomy?
Result of “overestimation” of
primary tumor category
Result of center-oriented accrual
(70% patients from two centers)
Discussion with regard to literature
Neoadjuvant chemotherapy has an effect on gastric cancer
• Who benefits the most?
• nCtx appropriate for all patients or only a subgroup?
• Will we be able to predict response in the future?
• Are there predicitive molecular tools or imaging tools available?
• Are there new and active regimen including biologicals?
• What is the role of radiochemotherapy?
Neoadjuvant Chemotherapy has a relevant toxicity
• Which is the least toxic but most effective regimen?
Influence of surgeon’s experience on survival
• Japanese training and expertise as one of our important aims
International cooperation in gastric research
• Support joint trials (MRC ST03, CRITICS, etc)
• Cooperation for future trials
Acknowledgements
Ulrich Fink

Professor emeritus

Medical Oncologist

Visionary Physician

Empathetic Personality
Thanks to Patients and Participating Centers
W. Bechstein, Frankfurt; P. Reichardt, Bad Saarow;
CF. Eisenberger, Düsseldorf; A. Hoelscher, Koeln;
H. Wilke, Essen;
Munich Center : Katja Ott, Sonja Gillen,
Maria Leibl and Rana Tabash;
EORTC Headquarters
M.A. Lentz, B. Meulemans, M.L. Couvreur, U. Bethe,
J Welch, B. Hasan, M. Mauer, L. Collette